"Tip: 2 out of your 500 calibration samples are associated to predicted values more extreme than the ones present in the isoscape. This corresponds to extrapolation during the calibration step. If the the proportion becomes too large, it could imped the reliability of your assignments. In such cases, you should perhaps rethink the design of your isoscape and/or collect more callibration data within the expected range to avoid any problem."
"Tip: 2 out of your 500 calibration samples are associated to predicted values more extreme than the ones present in the isoscape. This corresponds to extrapolation during the calibration step. If the the proportion becomes too large, it could imped the reliability of your assignments. In such cases, you should perhaps rethink the design of your isoscape and/or collect more callibration data within the expected range to avoid any problem."